Every great institution has its luminaries -- men and women whose contributions are of such brilliance
and influence as to bring credit to the institution as well as the individual. In April of this year,
IBM Research lost one of its brightest stars, Rolf Landauer, whose death from brain cancer ended a 47-year-long career with IBM.
Born February 4, 1927, to a Jewish
family in Stuttgart, Germany, Landauer came to New York in 1934.
After earning his doctorate at Harvard in 1950, Landauer worked for two years
in a laboratory that later became part of NASA before joining IBM in 1952 at
its Poughkeepsie, New York, laboratory.
From 1962 until 1969, when he was appointed an IBM Fellow and returned full time
to research, Landauer served in several management positions, including IBM assistant
director of research. Among the accomplishments that can be credited to his managerial
influence was the decision to pursue a program in large-scale integration. That initiative
helped sway IBM to develop metal-oxide-semiconductor (MOS) technology and fostered research
that resulted in the invention of dynamic random access memory (DRAM) and the discovery of
two-dimensional electron gases. Landauer also initiated the project that led to the invention
of the semiconductor laser, a feat achieved in a virtual dead heat with scientists at General
Electric and MIT's Lincoln Laboratory.
As a physicist, Landauer was best known for his novel theory of electron transport and his pioneering
work in the physics of information. He based his approach to electrical conductance on the idea of
electron scattering. He formulated that notion in terms of the probability that electrons entering a
conductor will emerge at the far end rather than be returned to the entry point. Initially ignored
when first published in 1957, his ideas were rediscovered in 1980, and the expression "Landauer's
formula" came into widespread use.
Perhaps Landauer is best known for having elevated the physics of information processing to a serious
field of scientific study. His dedication to this field grew out of a profound belief that "information
is physical'' and cannot be understood in isolation from its physical embodiment. In 1961, Landauer reexamined
the accepted belief that each elementary information-processing operation consumes a minimum amount of energy,
comparable to the energy of a molecule at room temperature. He showed that only irreversible operations -- those
that discard information -- are subject to that limit, a concept now known as Landauer's principle. This later led to
the notion of reversible computing and new concepts for the design of energy-saving circuits for very low power
applications, as well as to the modern understanding of Maxwell's Demon. The same convictions led Landauer to be
very suspicious of mathematical ideas, such as the infinitely many digits of the number pi, which have no
way of being realized in the physical world.
Landauer was respected -- and sometimes feared -- for his outspoken criticism of what he regarded
(correctly, it often turned out) as unsound or overly fashionable ideas in science and technology.
Yet his combination of technological savvy and physical insight provided a potent basis of evaluating
new proposals. In recent years, he was known for his skepticism about the now fashionable field of
quantum computation. In particular he doubted that delicate quantum states could be protected from
decoherence and noise long enough to do any useful amount of computation. This criticism stimulated
others to invent quantum error correction techniques which, as Landauer himself acknowledged, have
significantly improved the prospects of quantum computing -- though not, he said, to the point
where he'd buy stock in it.
Charles H. Bennett is an IBM Fellow at the Thomas J. Watson Research Center and a pioneer in the field of
quantum information.
